The application relates to a system for collection and analy sis of healthcare related data.

Currently, individuals, government entities, private compa nies, universities, and research and development institutes contribute to increase of human knowledge. Information or data of this knowledge is then shared with the public. This data has also increased rapidly and immensely.

In the healthcare industry, massive digitization of patient medical records, proliferation of numerous patient medical im aging technologies, and the advent of patient deoxyribonucleic acid (DNA) sequencing for use as a part of standard clinical practices have generated a large amount of medical data.

These medical data can include text that are provided in vari ous forms, such as structured text, semi-structured text, and unstructured text. The medical data can also include video, audio, and image information. These medical data can be gener ated and collected on a multitude of different platforms, such as social media portals, Internet of Things (IoT) sensor net works, and health record systems.

At the same time, technologies for analysing and understanding of these data for a variety of purposes are also developing. These technologies can enhance efficiency and productivity of businesses. They also can assist people to better manage their health and wellness .

It is an objective of the application to provide an improved system for providing personalised patient healthcare and for directing research and development (R&D) efforts to improve patient healthcare using data analytics.

The application provides a personalized healthcare system that comprises a computer. The personalized healthcare system re fers to a computer implemented system for providing healthcare services for diagnosing health issues of patients and for providing treatments that are personalized or customized for the respective patients .

The computer includes a communication interface for receiving medical image data and corresponding medical diagnoses of dif ferent patients, corresponding medical treatment histories of the different patients, and medical literature from various sources, such as clinics, hospitals, and medical institutions or health statutory organizations. The medical image data re fers to, for an example, images of human body parts, human tissues, or human organs for use in clinical diagnosis. The medical literature refers to, for an example, scientific lit erature, which includes articles in journals and books.

The computer also includes a computing processor for executing instructions of a diagnosis and treatment software program.

This diagnosis and treatment software program includes a data management component. The data management component comprises a plurality of databases. The databases refer to a collection of information that is organized so that it can be easily ac cessed, managed and updated. These databases are intended for storing the medical image data and the corresponding medical diagnoses of different patients, the corresponding medical treatment histories of the different patients, and the medical literature, which are received from the various sources. The diagnosis and treatment software program also includes an analytics component, which includes a plurality of modular an alytics modules. These modular analytics modules can be put together in different combinations to perform different tasks or functions .

The plurality of modular analytics modules includes a data fu sion module for integrating the received medical image data and the corresponding medical diagnoses, the received corre sponding medical treatment histories, and the received medical literature to provide treatment plan datasets. In other words, the treatment plan datasets contain integrated data, which have been parsed, cleaned, transformed, and harmonized for im proved data quality.

The analytics modules also include an image analytic module for constructing a model of the medical image data and the corresponding medical diagnoses using a machine learning algo rithm and for providing or identifying a possible diagnosis of a health related issue of a patient using the model. The image analytic module can use computational statistics to analyse the received medical image data for making a prediction of a likely health related issue.

The analytics modules also include a medical treatment module for providing possible medical treatment options for the iden tified health related issue according to the treatment plan datasets .

The diagnosis and treatment software program further include a visualization component for receiving a query regarding the health related issue from a user. The diagnosis and treatment software program further includes a chat-bot module for receiving the query from the visualization component. The chat-bot module then provides a response to the query. The response is later transmitted to the user via the visualization component.

The chat-bot module can be adapted to process and comprehend the query using natural language processing. The natural lan guage processing refers to a software program, which includes algorithms with an ability to understand human language sen tences or natural language sentences. The algorithms can in terpret key elements of the natural language sentences, and then retrieve relevant information that corresponds to the query from at least one database. The algorithms later build at least one natural language sentence with the retrieved rel evant information for sending it to the user.

The query can be provided using either a text message or an audio message.

The computer can be provided by a cloud computing system. The cloud computing system comprises a network of computers. These computers provide a significant portion of hardware and soft ware support functions of outsource facility management ser vices . The cloud computing system also allows additional com puting resources to be easily added using techniques, such as load balancing, for handling surges in user demand. Multiple regional computing clusters can also be set-up to support dif ferent regional computing needs. The cloud computing system enables the supporting network computers to be located in an other location that is away from a user with a computing de vice. The computing device of the user can access the cloud computing system via a web application over the Internet. The application also provides a computer media that stores a diagnosis and treatment software program of a personalized healthcare system, which is described above. The computer me dia refers to data storage devices.

The application further provides a method for providing per sonalized healthcare for a patient. The method comprises a step of receiving medical image data and corresponding medical diagnoses of different patients, corresponding medical treat ment histories of the different patients, and medical litera ture and a step of integrating the received medical image data and the corresponding medical diagnoses, the received corre sponding medical treatment histories, and the received medical literature to provide treatment plan datasets. The method fur ther includes a step of constructing a model of the medical image data and the corresponding medical diagnoses using a ma chine learning algorithm and a step of providing a possible diagnosis of a health related issue of a patient using the model. The method also includes a step of providing possible medical treatment options for the health related issue accord ing to the treatment plan datasets.

The method can further include a step of receiving a query in at least one natural language sentence for identifying a healthcare related issue from a user and a step of processing and comprehending the query using natural language processing. A step of retrieving relevant information that corresponds to the query from at least one database can then be executed. The method further includes a step of building at least one natu ral language sentence with the retrieved relevant information, and a step of sending at least one natural language sentence containing the relevant information to the user. The method can further include a step of selecting a possible medical treatment option that is aligned with an insurance policy of the patient.

Fig . 1 illustrates an improved healthcare system for

providing personalised patient healthcare and for directing research and development (R&D) efforts to improve patient healthcare,

Fig . 2 illustrates a block diagram of a data analytic and modelling software module of a data analytics device of the system of Fig. 1,

Fig . 3 illustrates a sandbox testing environment of the data analytic and modelling software module of Fig. 2,

Fig . 4 illustrates a flow chart of a method for operating the system of Fig. 1,

Fig . 5 illustrates a clinical workflow of diagnosis and treatment of a breast cancer patient, the diagnosis and the treatment being guided by the system of Fig. 1 , and

Fig . 6 illustrates a clinical workflow of diagnosis and treatment of a ROP patient, the diagnosis and the treatment being guided by the system of Fig. 1.

In the following description, details are provided to de scribe embodiments of the application. It shall be apparent to one skilled in the art, however, that the embodiments may be practiced without such details .

Some parts of the embodiment have similar parts. The similar parts may have the same names or similar part numbers. The de scription of one similar part also applies by reference to an other similar part, where appropriate, thereby reducing repe tition of text without limiting the disclosure. Fig. 1 shows a block diagram of an improved healthcare system 1. The healthcare system 1 includes a plurality of data sources 10, a data analytics device 3, multiple information visualization devices 7, and a communication network 13.

The data analytics device 3 is communicatively connected to the information visualization devices 7 and to the data sources 10 via the communication network 13.

In detail, the data sources 10 includes a plurality of data bases for storing healthcare related data. The data can be provided in a form of structured data and unstructured data. The data can include text and/or images. Examples of the data include laboratory genomic data, medical prescription notes, clinical case notes, medical images such as mammograms using X-rays and images using ultrasound imaging or magnetic reso nance imaging (MRI), pharmacy data, drug toxicity data, clini cal pathways, and medical literatures.

The information visualization devices 7 include smart phones, tablets, laptop computers, and/or computers. The smart phone includes a mobile phone that includes a computer. The mobile phone also has a touchscreen interface, an Internet access, and an operating system capable of running software applica tions .

The data analytics device 3 includes a computing processor 31, a memory module 34, and a communication interface 38. The com puting processor 31 is electrically connected to the memory module 34 and to the communication interface 38. The communi cation interface 38 is communicatively connected to the commu nication network 13. The memory module 34 stores an operating system 36 and software applications. The computing processor 31 acts to execute instructions, which are provided by the software applications .

The communication network 13 includes the Internet, an intra net, a wide-area network (WAN) , a local area network (LAN) , and/or a wireless network.

Regarding the data analytics device 3, the software applica tions include a data analytic and modelling software module 40, which is shown in Fig. 2. The data analytic and modelling software module 40 includes a data management software compo nent 50, an analytics software component 60, and a visualiza tion software component 70. The data management software com ponent 50 is also called a data management software layer. The analytics software component 60 is also called an analytics software layer. The visualization software component 70 is also called a visualization software layer.

Referring to the data management software component 50, it in cludes a patient clinical data mart 53 and a research and de velopment data mart 56.

The patient clinical data mart 53 includes a medical imaging database, a patient medical prescription database, and a la boratory data genomics database as well as a database to store patient clinical case notes and patient hospitalization rec ords. As an example, the medical imaging database can store mammogram images .

The research-and-development data mart 56 includes a medical literature database, a drug toxicity database, a clinical pathway database, and knowledge of the crowd database. Each database is provided as one or more collections of data. The database act as a central repository of data, which are received from one or more disparate data sources 10. The data can be provided in multiple data types of different modality and can be in different formats, such as in text and images. The data can also be longitudinal in that it extends over pa tient health history or medical record. The data are often parsed, cleaned, transformed, and harmonized before the data are used. The data can also anonymized, wherein all infor mation for identifying patient are removed from the data to prevent identification of patient.

These patient clinical data mart 53 and the research and de velopment data mart 56 can be stored in computer servers, cloud instances, personal computers, laptops, and/or portable devices .

Referring to the analytics software component 60, it includes a library comprising a plurality of analytics modules for healthcare and biomedical applications .

The analytics modules include a data fusion module 62, an im age module 64, a risk profiling module 66, and a descriptive statistics module 68.

In detail, the data fusion module 62 is used for gathering multiple heterogeneous data that are extracted from the data bases of the data management software component 50, for inte grating or harmonizing the multiple heterogeneous data, and for storing the harmonized data into the databases.

The image module 64 is used for automating assessments or di agnosis of medical images to identify a likely health related issue, such as a disease. The risk profiling module 66 is used for providing patient segmentation and patient risk stratification according to the harmonized data.

The descriptive statistics module 68 uses an artificial intel ligence algorithm, such as machine learning, to provide in sights of disease progression of individual patients for im proving patient care.

The descriptive statistics module 68 also provides insights of effectiveness of various treatments to private or public re search organizations, thereby allowing the research organiza tions to streamline or direct their research and development (R&D) activities for improving clinical care.

One example of the insight relates to treatment plans for pa tients, wherein the treatment plans are personalized according to a patient profile, patient genome information, diagnostic test results of the patient, and past treatments of similar patients .

Referring to the visualization software component 70, it in cludes an interactive interface module 73 and a plurality of visualization modules 76.

The interactive interface module 73 receives inputs, which are provided from a user to the information visualization device 7. The interface module 73 then transmits the user inputs to the analytics software component 60. The user can refer to a healthcare professional such as a doctor, an oncologist, a ra diologist, and a genomic practitioner, to an insurance pro vider, and to a staff of a pharmaceutical company. The analytics software component 60 then generates information according to the user inputs. The visualization module 76 later receives the generated information from the analytics software component 60 and then displays the generated infor mation on the information visualization device 7 for viewing by the user.

In a general sense, the data analytic and modelling software module 40 of the data analytic device 3 is scalable.

The data analytic and modelling software module 40 contains components or modules that can be added or removed, thereby allowing for dynamic readjustment and customization of the data analytic and modelling software module 40. Modules or components of the data analytic and modelling software module 40 can change as data and/or data analytics strategy evolve for individual user.

For an example, a clinical partner wants to incorporate an ad ditional customized database into the data analytics device 3. The data analytics device 3 is configured for management and automated grading of mammograms and for providing patient treatment information. The data analytics device 3 then allows the data management software component 50 to be scaled-up, by including the additional customized database.

The modular nature of the data analytic and modelling software module 40 also allows users to add components to the library of analytics modules for enhancing patient care. The compo nents can include customized modules or modules of existing libraries .

In a special embodiment, as seen in Fig. 3, the data analytic and modelling software module 40 includes a sandbox testing environment 63 for allowing users to test and optimize or im prove solutions or results that are provided by the data ana lytics device 3. The sandbox testing environment 63 provides a computing space for users to add and remove databases, and modules or components for evaluation without affecting other functions of the data analytics device 3.

The healthcare system 1 includes a cryptographic security sys tem for meeting security requirements, such as health insur ance portability and accountability act (HIPAA) , and other healthcare industry regulations .

The cryptographic security system acts to protect data stored in the patient clinical data mart and the research and devel opment data mart of the data management software component 50.

As an example, database architectures, such as non-relational (NoSQL) databases, of the patient clinical data mart and the research and development data mart are provided with encryp tion and/or password to protect and to safeguard data, which are stored in these database architectures .

In another example, the analytics modules are provided with differential privacy and encryption to ensure data security and privacy when performing data analytics .

The cryptographic security system acts to protect data/in sights that are being transmitted over the communication net work 13 using various data security features, such as encryp tion protocol. The encryption protocol can refer to Transport Layer Security (TLS) and/or Secure Sockets Layer (SSL) proto cols . The cryptographic security system also provides dedicated ac cess controls for controlling user-rights and for controlling access to different parts of the healthcare system 1, namely the data analytics device 3 and the information visualization devices 7.

The cryptographic security system also provides data integrity controls for protecting encryption keys that are used by cryp tographic security system.

Real-time security and compliance monitoring are also carried out to detect anomalous activities using security information and event management (SIEM) software products and services like open source security information management (OSSIM) sys tem.

In summary, the healthcare system 1 provides several benefits .

The healthcare system 1 provides flexibility, scalability, and security. In detail, the healthcare system 1 contains compo nents that are modular, thereby allowing users to add and re move components for meeting different needs of the users . The healthcare system 1 is also scalable. The modular nature of the data analytics device 3 also allows dynamic adjustment and customisation of databases, the analytics modules, and/or the visualization modules according to changing need of the user. The user can add or remove any components that are provided in the data analytics device 3 as and when they need. The healthcare system 1 is also secure for preventing unauthorised access of the healthcare system 1.

The data sources 10 can also include patient medical databases of multiple healthcare service providers, wherein the healthcare service providers collaborate with insurance pro viders to provide medical services for insured persons who have purchased travel insurance policies from their respective insurance providers .

These insured persons can provide consent to allow the im proved healthcare system 1 to retrieve their medical infor mation from the databases of these healthcare service provid ers when the travel insurance policies of the insured persons are in force .

When an insured person has a healthcare emergency during an oversea trip, the insured person or his family member can no tify his insurance provider. The insurance provider can then request the improved healthcare system 1 to retrieve medical records of the corresponding insured person from his

healthcare service provider and later transmit the retrieved medical records to an oversea healthcare service provider that is taking care of the healthcare emergency of the insured per son .

This arrangement allows the overseas healthcare service pro vider to obtain relevant medical information quickly, thereby avoiding unnecessary medical tests, which are needed to treat the patient. This can then save time and cost in treating the insured person.

Fig. 4 shows a flow chart 80 of a method of providing an on cology healthcare to a patient, which uses machine learning technologies to process large patient data relating to breast cancer for diagnosing possible breast cancer progression of the patient and for providing possible personalized treatments for the patient. The flow chart 80 includes a step 82 of providing possible di agnosis .

In detail, the computing processor 31 executes instructions to gather different medical diagnostic images of human breasts with corresponding diagnosis of various patients from a plu rality of clinics, hospitals, and medical institutions or health statutory organizations, in a step 84. Examples of the medical diagnostic images include mammograms, ultrasound im ages, and MRI images.

After this, the computing processor 31 executes instructions of an algorithm of a data fusion module for integrating the above-mentioned data to form human breast image datasets with a predetermined data structure, in a step 86.

In a next step 88, the computing processor 31 executes in structions to augment the human breast image datasets by gen erating additional image datasets. The data augmentation, as an example, includes scaling and rotation of the medical diag nostic images of human breasts .

The computing processor 31 then executes instructions to con struct a breast cancer image diagnostic model from the human breast image datasets with corresponding diagnosis using ma chine learning techniques, in a step 90.

The image diagnostic model can be later used to provide a sug gested diagnosis of a medical diagnostic image of a new pa tient. The suggested diagnosis can relate to possible identi fication of breast tumours as well as possible growth stage of the breast tumours. The suggested diagnosis is later stored in one or more databases of the oncology healthcare system 1. Following this, a step 92 of providing possible treatments is performed .

In detail, the computing processor 31 executes instructions of an algorithm of a data fusion module for gathering various medical image diagnoses, liquid biopsy test results, breast biopsy pathology reports, and patient case histories of vari ous patients with various profiles from various clinics, hos pitals, and medical institutions or health statutory organiza tions, in a step 94. The patient case histories include treat ments received by the patients .

After this, these data are integrated to form human breast treatment datasets, in a step 96.

The data from various medical publications relating to breast cancer can also be integrated into the human breast treatment datasets. The medical publications include medical literatures and medical care standards or treatment guidelines.

The treatment datasets can later be used to provide possible treatment plans for a new patient, in a step 98, wherein each possible treatment plan is personalized or customized for the new patient according to medical history and test results of other patients with profile similar to the new patient and ac cording to relevant medical literature. The possible treatment plans are then stored in one or more databases of the oncology healthcare system 1.

The oncology healthcare system 1 can further include a chat bot module, which is provided in an analytics software compo nent 60 of the oncology healthcare system 1. The chat-bot mod ule refers to a software program that provides a method for providing a conversation between the oncology healthcare sys tem 1 and a user.

In detail, a user provides a query in the form of one or more natural language sentences to a chat-bot interface of an in formation visualization device 7. The sentences can be in a text format or an audio format .

An interactive interface module 73 of a visualization software component 70 then receives the query and later sends to the chat-bot module.

The chat-bot module afterward receives the query and then pro cesses and comprehends the query using natural language pro cessing techniques.

After this, the chat-bot module builds one or more natural language sentences with the relevant information in a text format or an audio format.

The natural language sentences that contain the relevant in formation are later sent to the user.

In one implementation, the chat-bot module is configured to interface with databases, which store medical information of various patients, for access by clinicians only. This clini cian chat-bot module can allow a doctor to issue a query about the patient to retrieve medical information of a patient eas ily for display on his information visualization device 7 when the doctor delivers healthcare services to the patient at a point of care.

This clinician chat-bot module can also allow, for an example, an oncologist to order a targeted therapy drug for a cancer patient using natural language sentence request. This will make the drug ordering process efficient and simple, thereby removing barriers that may hinder the oncologist from procur ing the targeted therapy drug.

In another implementation, the chat-bot module is configured to interface with databases, which store, for examples, pa tient treatment plans and information relating to nutrition and lifestyle guidance for patients to access. This patient chat-bot module can allow patients retrieve their respective treatment plans for viewing. The patients can also receive nu trition and lifestyle guidance for improving their health. Furthermore, the patients can also provide feedback about their treatment progresses, to their respective doctors, via a chat-bot interface.

Moreover, the oncology healthcare system 1 is also configured to provide clinical workflow steps for diagnosis and treatment of the breast cancer, which are described below.

Fig. 5 shows a clinical workflow 100 of diagnosis and treat ment for a breast cancer patient, wherein the oncology healthcare system 1 is used to provide possible diagnoses and possible treatments .

The workflow 100 shows a first step 110 of a patient detecting a breast lump during a physical or medical examination or de tecting indications of cancer cells in a blood sample during a regular liquid biopsy screening.

The patient then consults with a physician, in a step 120. The physician later inputs a query to the oncology healthcare system 1, via an information visualization device 7, to re trieve electronic medical records of the patient.

In a next step 130, the physician uses the retrieved medical records to assist him to provide a clinical assessment of the patient .

The physician then updates the patient electronic medical rec ords in the oncology healthcare system 1 with his clinical as sessment of the patient.

The oncology healthcare system 1 afterward analyses the clini cal assessment and suggests that the patient undergoes a mam mogram, in a step 140. The oncology healthcare system 1 also identifies possible clinics that can provide mammography ser vices with their corresponding waiting periods.

A staff of the physician later selects a clinic and also ar ranges a mammogram appointment with the selected clinic for the patient using the oncology healthcare system 1.

A radiologist of the selected clinic then performs a mammogram for the patient and later stores images of the mammogram in the databases of the oncology healthcare system 1.

In a subsequent step 150, the oncology healthcare system 1 analyses the mammogram images using the image diagnostic model to provide possible diagnosis. The diagnosis can include a recommendation for further diagnostic tests, such as an ultra sound scan or a MRI scan. The diagnosis can also include a recommendation for a breast biopsy. After this, the radiologist reviews the mammogram images and the possible diagnosis, which is generated by the oncology healthcare system 1. The radiologist then provides a final di agnosis of the mammogram images. The possible diagnosis acts to assist and guide the radiologist to provide the final diag nosis, wherein the radiologist decides whether to take up these suggestions .

The oncology healthcare system 1 also identifies clinics that can perform the further diagnostic tests and also allows a staff to arrange appointments with the selected clinics for the patient.

Healthcare professionals of the selected clinics can later perform the recommended diagnostic tests for the patient and can provide information about results of the diagnostic tests to the oncology healthcare system 1.

The oncology healthcare system 1 later performs a diagnosis of the patient regarding disease progression based on the diag nostic test results .

The oncology healthcare system 1 also suggests next diagnostic steps according to the human breast treatment datasets and also identifies possible medical oncologists for the patient, accord to the patient profile and past medical case history of the patient. The physician later selects a medical oncologist who specialises in breast cancer for the patient, in a step 160.

In a next step 170, the selected medical oncologist reviews the possible treatment plans and selects next steps, such as obtaining metastases scan on high risk body areas, obtaining a multiplex i munohistochemistry scan, and/or performing genome testing for obtaining patient genomic sequencing.

Results of the scan and of the genome testing are afterward transmitted to the oncology healthcare system 1 for storing in respective case files of the patient.

The healthcare system 1 then generates possible treatment plans, according to the stored date, using the treatment da tasets, for the medical oncologist to review, in a step 180.

The treatment plans are personalized or customized according to the scan results and to the genome testing results together with other patient test results and with other patient related data, such as the patient profile and patient past medical case history. The treatment plans are also formulated accord ing to clinical literature, medical guidelines, and/or past treatments of other similar patients, for producing the best outcomes. The treatment plans may include lumpectomy or mas tectomy, chemotherapy, radiotherapy, and hormone therapy.

The medical oncologist then reviews the possible personalized treatment plans with supporting medical evidences and with in surance coverage of the patient, which are provided the healthcare system 1.

The medical oncologist can later discuss the possible person alized treatment plans with the patient and then select a treatment plan that is aligned with the patient insurance cov erage .

During the course of treating the breast cancer, the medical oncologist can also use the oncology healthcare system 1 to arrange schedules for performing liquid biopsy tests and/or breast biopsies for monitoring the patient's response to the treatments .

Results of the liquid biopsy tests and pathology reports of the breast biopsies are also stored in the oncology healthcare system 1 to assess disease progression and response of the pa tient to the treatments performed so far.

The oncology healthcare system 1 later generates an analysis of these test results and these pathology reports .

The oncology healthcare system 1 may suggest changes to the treatment plan, if any, based on the assessment. For an exam ple, the system may select and recommend a clinical trial op tion for the patient.

The medical oncologist then reviews the suggested changes to the treatment plan and can adjust the treatment plan accord ingly .

Healthcare professionals, such as nurses, can also use the on cology healthcare system 1 to discuss scheduling of the treat ment with the patient to optimise or improve the treatment schedules .

The oncology healthcare system 1 can also automatically place orders of required drugs and medical services ahead of sched ule .

When the cancer is no longer detected, the medical oncologist can schedule the patient for follow-up monitoring, which can extend over several years, in a step 190. During this time, the patient can be subjected to annual mammography, clinical review every 3 to 6 months for a period of 5 years, monthly breast self-examination, and nutrition and lifestyle guidance.

The oncology healthcare system 1 can also monitor the follow up monitoring and issue automated reminders to the patient re garding the follow-up monitoring.

The oncology healthcare system 1 can also provide nutrition and lifestyle guidance to the patient based on analysis of case histories of similar patient profiles.

The outcome of the follow-up monitoring is later stored in the databases of the healthcare system 1 to allow the oncologist or his staff to monitor and track the health condition of the patient .

In summary, the oncology healthcare system 1 provides several features .

For patients, the oncology healthcare system 1 is patient-cen tric, being directed to provide best or improved treatment outcome for the patient. The oncology healthcare system 1 pro vides treatment plan options that are data-backed for a pa tient to select a treatment plan that is targeted or custom ized to suit the patient. The patient can then choose the treatment plan according to cost, namely patient financial ability and patient insurance plan. The patient can then have higher trust in their clinicians and in treatment and recovery processes. The patients can also recover faster with fewer complications, allowing the patient to resume their former life activity.

For doctors, the oncology healthcare system 1 provides a pa tient-specific decision support. The decision support provides doctors with patient-specific information that is integrated and analyzed for presenting at a point of care so that the doctors can deliver consistent and high-quality cancer care. The oncology healthcare system 1 also provides possible analy sis with possible treatments for doctors to view in order to produce the best possible outcome for patients with similar profiles and for doctors to adopt best practices. The oncology healthcare system 1 also provides consistent clinical workflow for doctors .

The oncology healthcare system 1 also can enable the doctors to treat their patients faster, and at lower cost, which con tributes to their job satisfaction and also to enhance their reputation .

The oncology healthcare system 1 also enables the doctors to view respective tier of insurance plan of the patients, so that they can discuss various treatment plans and correspond ing costs with the patients for aligning with their insurance plans .

In circumstances where financial assistance programs are available to patients, the oncology healthcare system 1 can also be further adapted to inform the about this, in order to enroll patients in the respective patient financial assistance programs. This not only eases the financial burden of the pa tients, but also saves time in the enrollment process.

Moreover, the oncology healthcare system 1 allows doctors to be more efficient so that they can join a pool of doctors who are approved by insurance providers .

The oncology healthcare system 1 also helps doctors to contact with other approved doctors for discussing about further treatment plans of their patient. In effect, the healthcare system 1 acts to enlarge the network of the doctors and also provides the doctors with a list of approved doctors who are suitable for their patients.

For insurance providers, the oncology healthcare system 1 al lows insurance providers to have a pool of doctors partnering and collaborating with the insurance providers.

There is often a high demand for doctors to join this pool of collaborating doctors because the insurance providers can pro vide a constant flow of patients for these collaborating doc tors, wherein these patients hold medical insurance policies with the insurance providers .

The oncology healthcare system 1 can also enable the insurance providers to rank these collaborating doctors.

The oncology healthcare system 1 is able, over a period of time, to gather statistics regarding costs and effectiveness, as well as other parameters of these collaborating doctors.

The cost can relate to number of claims and to claim amount of each patient of each of the collaborating doctor.

The insurance providers can then remove least effective or most expensive doctors from the pool of the collaborating doc tors, thereby reducing an overall number of claims and reduc ing the amount of the claims, which in turn lowers cost of in surance and reduces processing time of claims.

In addition, the oncology healthcare system 1 can also allow the insurance providers to formulate new insurance products with new pricing models, such as offering new personalized in surance products that can be aligned to health risk profile of a person.

Fig. 6 shows a flow chart 200 of a method for diagnosing pos sible ROP using this eye vision healthcare system 1.

Herein, the improved healthcare system 1 is configured to pro vide a method for eye vision care using machine learning tech nologies to process video images of babies' eyes for diagnos ing possible retinopathy of prematurity (ROP) and identifica tion of a possible stage of ROP.

Retinopathy of prematurity is an eye illness that happens to premature babies. The premature babies refer to babies being born after a gestation of around 36 to 37 weeks or shorter. A normal pregnancy lasts for about 40 weeks. Consequently, the premature babies have less time to develop in the womb of their mothers, causing them to be more susceptible to health complications and defects. Some of the problems that may af fect the premature babies include vision and hearing impair ment. This is because final stages of vision and hearing de velopment of the baby occur in last few weeks of pregnancy.

The premature babies, therefore, have higher risks in develop ing eye and ear problems .

Many of these eye problems stem from abnormal development of blood vessels of the eyes, which can lead to vision impair ment. While eyes of the babies may look normal, but they do not respond to moving objects or changes in light intensity. These abnormalities may be signs of a vision problem or an eye defect . Retinopathy of prematurity can lead to a serious complication of prematurity. ROP is a significant cause of preventable blindness in children and it accounts for about six to eight een percent of all childhood blindness in developed countries . This figure is probably higher in undeveloped countries.

Most babies with ROP disease show symptoms only a few weeks after they are born. By then, the babies may already have left the hospitals. A large population in developing countries lives in rural areas and they often lack resources to go back to urban hospitals, when the babies start showing these symp toms .

Babies who are born prematurely often need intensive neonatal care. Oxygen therapy is often used due to premature develop ment of the babies' lungs. However, this may result in disease of the eye. It is thought that such disease is caused by dis organized growth of retinal blood vessels. The retina may be come damaged if abnormal blood vessels begin to swell and to leak blood. When this happens, the retina can detach from the eyeball, leading to a vision impairment.

The retinopathy of prematurity comprises 5 stages. In stage 1, an affected eye shows a demarcation line. In stage 2, the de marcation line has a ridge and/or popcorn appearing on the ridge. In stage 3, new blood vessels - neovascularization - are developed on the demarcation line and this can lead to ex tra-retinal proliferation of tissue. The ridge may bleed and cause traction on the ridge. The stage 3 is a critical stage in development of ROP as it signifies that retinal ischaemia is significant enough to cause development of new vessels for supplying oxygen to the retina. However, since retinal vessels are not there in the avascular retina in the periphery, the new vessels grow in a futile attempt to try to re-vascularise the area. In stage 4, retinal detachment starts to occur due to the traction of the fibrous and vascular tissue over the ridge. The stage 4 includes a stage 4A, wherein the retina, without its macula, detaches from the eyeball and a stage 4B, wherein the retina with its macula detach from the eyeball. In stage 5, which is the final stage of ROP, the retina com pletely detaches from the eyeball, causing vision impairment.

ROP can be mild and may resolve spontaneously. But it can also lead to blindness in serious cases. ROP can also cause other complications later in life of the baby. Examples of these complications include crossed eyes, near-sightedness, far sightedness, lazy eye, and glaucoma. It is hence important to screen infants for ROP.

ROP can be screened or diagnosed by ophthalmologic examina tion. The ophthalmologic examination can be done using a bin ocular indirect ophthalmoscopy (BIO) . This method, however, often does not permit adequate assessment of ROP in retinal periphery. Moreover, appropriate pain-relieving steps, such as local anaesthetic eye drop and oral sucrose, need be used. The ophthalmologic examination can also be performed using a retcam, which refers to a wide angle digital paediatric reti nal imaging equipment. This method avoids stress and expertise of BIO examination and indentation but is as specific and sen sitive as BIO. This method is useful for diagnosis and tele medicine .

The ophthalmologic examination is often labour intensive, and the examination procedure is also often uncomfortable for the premature babies. Unlike adults who can receive doctor's in struction and act accordingly when they receive ophthalmologic examination using an ophthalmoscopy test to see inside of the eye, the premature babies are often active and cannot act ac cording to doctor's instructions. Furthermore, light is not suitable to be shone directly into their eyes, as this can lead to possible blindness.

The improved healthcare system 1 for eye vision care can re duce dependence on skilled medical personnel for ROP screening and for reducing their workload. This can address a problem of the number of trained doctors being insufficient to take care of the many babies in various developing and under developed countries. The use of the eye vision care system can also shorten the time for detecting the ROP symptoms to allow doc tors performing treatment on patients early.

The method includes a first step 210 of a computing processor 31 executing instructions to gather different human eye raw image data, which includes ROP image data with corresponding diagnosis that include ROP stage data, from a plurality of clinics, hospitals, and medical institutions. The human eye raw image data can include digital and/or video eye images that show eye features, such as eye demarcation line, eye de marcation line with ridge, popcorn appearance on the ridge, ridge with extraretinal proliferation, and partial or total retinal detachment.

The computing processor 31 then executes instructions of an algorithm of a data fusion module for integrating the eye raw image data to form eye datasets, in a step 220.

The computing processor 31 later executes instructions to aug ment the eye datasets by generating additional eye datasets, in a step 230. The computing processor 31 then executes instructions to con struct a ROP model from the eye dataset, using a machine learning algorithm, in a step 240.

This ROP model can later be used for providing a suggested di agnosis of eye video images of a new baby. The suggested diag nosis can also include identification of ROP as well as stage of the ROP.

The improved healthcare system 1 for eye vision care can also include a chat-bot module for providing a conversation between the improved healthcare system 1 and a user. This chat-bot module is similar to the chat-bot module of the oncology healthcare system 1, which is described above.

In a general sense, the improved healthcare system 1 can also be configured to provide possible diagnosis and possible treatment of other health related issues, such as diabetes, wherein the possible treatment can include personalized diet plans for diabetes patients.

In summary, the improved healthcare system 1 provides compo nents that are modular for allowing users to add and remove components for meeting different needs of the users. This also allows dynamic adjustment and customisation of databases, the analytics modules, and/or the visualization modules according to changing need of the user.

The embodiments can also be described with the following lists of features or elements being organized into an item list. The respective combinations of features, which are disclosed in the item list, are regarded as independent subject matter, re spectively, that can also be combined with other features of the application. A personalized healthcare system comprising a computer, the computer comprising

a communication interface for receiving medical image data and corresponding medical diagnoses of different patients, corresponding medical treatment histories of the different patients, and medical literature, and

a computing processor for executing instruc tions of a diagnosis and treatment software program, the diagnosis and treatment software program comprising

a data management component comprising a plurality of databases for storing the medical image data and the corresponding medical diagnoses, the corresponding medi cal treatment histories, and the medical literature ,

an analytics component comprising a plu rality of modular analytics modules, the plurality of modular analytics modules comprising

a data fusion module for integrating the medical image data, the corre sponding medical diagnoses, the cor responding medical treatment histo ries, and the medical literature to provide treatment plan datasets, an image analytic module for con structing a model of the medical im age data and the corresponding medi cal diagnoses using a machine learn ing algorithm and for providing a possible diagnosis of a health re lated issue of a patient using the model ,

a medical treatment module for providing possible medical treatment options for the health related issue according to the treatment plan da tasets,

a visualization component for receiving a query regarding the health related issue from a user, wherein the diagnosis and treatment software program fur ther comprises a chat-bot module for receiving the query and for providing a response to the query for transmit ting to the user via the visualization component.

2. The personalized healthcare system according to item 1, wherein

the chat-bot module is adapted to process and comprehend the query using natural language processing, to retrieve relevant information that corresponds to the query from at least one database, and to build at least one natural language sentence with the retrieved relevant information for sending it to the user.

3. The personalized healthcare system according to item 1 or 2, wherein

the query is provided using a text message.

4. The personalized healthcare system according to item 1 or 2, wherein

the query is provided using an audio message. 5. The personalized healthcare system according to one of items 1 to 4, wherein the computer is provided by a cloud computing system.

6. A computer media storing a diagnosis and treatment soft ware program of a personalized healthcare system of one of items 1 to 4.

7. A method for providing personalized healthcare for a pa tient, the method comprising

receiving medical image data and corresponding medical diagnoses of different patients, corre sponding medical treatment histories of the different patients, and medical literature, integrating the medical image data and the cor responding medical diagnoses, the corresponding medical treatment histories, and the medical literature to provide treatment plan datasets, constructing a model of the medical image data and the corresponding medical diagnoses using a machine learning algorithm,

providing a possible diagnosis of a health re lated issue of a patient using the model, and providing possible medical treatment options for the health related issue according to the treatment plan datasets.

8. The method for providing personalized healthcare for a patient according to item 7, the method further compris ing

receiving a query in at least one natural language sentence for identifying a healthcare related issue from a user, processing and comprehending the query using natural language processing,

retrieving relevant information that corresponds to the query from at least one database,

building at least one natural language sentence with the retrieved relevant information, and

sending at least one natural language sentence con taining the relevant information to the user.

9. The method for providing personalized healthcare for a patient according to item 7 or 8 , the method further com prising

selecting a possible medical treatment option that is aligned with an insurance policy of the patient.

Although the above description contains much specificity, this should not be construed as limiting the scope of the embodi ments but merely providing illustration of the foreseeable em bodiments. The above-stated advantages of the embodiments should not be construed especially as limiting the scope of the embodiments but merely to explain possible achievements if the described embodiments are put into practice. Thus, the scope of the embodiments should be determined by the claims and their equivalents, rather than by the examples given.

REFERENCE NUMBERS

1 improved healthcare system

3 data analytics device

7 information visualization devices

10 data sources

13 computer/communication network

31 computing processor

34 memory module

36 operating system

38 communication interface

40 data analytic and modelling software module

50 data management software component

53 patient clinical data mart

56 research and development data mart

60 analytics software component

62 data fusion module

63 sandbox testing environment

64 image module

66 risk profiling module

68 descriptive statistics module

70 visualization software component

73 interactive interface module

76 visualization modules 76

80 flow chart

82 step

84 step

86 step

88 step

90 step

92 step

94 step

96 step

98 step

100 flow chart 110 step

120 step

130 step

140 step 150 step

160 step 170 step 180 step 190 step 200 flow chart

210 step 220 step 230 step 240 step